Capacitor Value affirmation, 19C: ME needs EE Help!


EE types

I am trying to revive a 29C that had bad corrosion. Two obvious replacements are the two capacitors marked



I'm running out of patience looking around the web trying to decode. The first one is 1000 pf and the second is what??? Right?

Edited: 22 Sept 2012, 9:43 p.m.



I once found this page about capacitors and I compared the codes expressed in the tables right after 'Capacitor Codes' (four clicks from the page top) to what I red with a capacitance meter: they all matched.

Hope this helps.


Luiz (Brazil)


These are interesting numbers. Capacitor values are ususally marked in 2 different ways: A three number code designating pF (i.e. value-value-exponent e.g. 473 = 47000 pF) or just a number designating uF (e.g. 0.1 = 0.1uF or 100nF). The letter after the number is usually the torlerance (in this case K=10%).

I your case it is possible that the first one is 0.001uF = 1nF = 1000pF as you have already deduced, and 1nF 1000V is not an unreasonable size. However, according to that the second one would be 3900uF 250V, that is a very large capacitor - especially 35 years ago.

Can you supply some good resolution pictures of these capacitors? Do you have circuit diagrams to try and trace these capacitors? (Apparently the 19c/29c circuit diagrams are on the HPCC CD).


Thanks Bart,

I have this CD. I thought the diagram must be somewhere. I will look and see if it can solve the mystery.


I think the voltage could be pretty much ignored - and 3900uF 250V sounds way out of proportion for this calculator (I doubt that it would fit into the calc <g>) , I'd rather assume 25V - 3900uF would otherwise make sense as a buffer capacitor for the printer. (The small pF and nF types often come as 400V and up types due to their construction, but not the electrolytic/tantal types with large capacities, those really grow in size when specified for higher voltage...)


The value in question is 3900 pf. Yes, it really is rated at 250 volts. The .001 uf (1000pf) is indeed rated at 1000v and was a common part back when the machines were built.

Unless the leads are physically damaged, there really is no need to replace them. They are high quality ceramic disc capacitors and they are pretty much impervious to corrosion.

The 3900pf cap is important as it establishes the frequency of operation of the switch mode power supply. The .001 uf is in the printer circuit and could be ignored until you get the calculator portion operational.

I would be careful when working on the board in question, it is very thin and is multi-layered (most likely three). A good desoldering station is a must. A hand held solder sucker or desoldering braid may cause more problems than you solve.

Edited: 24 Sept 2012, 10:08 a.m.


Thanks for the clarification, Randy. Interesting that it is not labelled as 392K - as would normally be the case for that value (even in the seventies).


In the seventies, it was not common for US component manufacturers to use shorthand numbering systems, since there was (at the time) lots of area for marking. Specialty resistors used the R,K and M decimal place markers but that was about it. With most caps, we pretty much assumed uf if the number began with a decimal point (or the better 0. notation), pf for whole numbers. It was not the best but that was the convention. To put that in the perspective of the time, we had "issues" adopting the metric system too...


This was probably one of the more favorable sides of the calculator wars: japanese manufacturers switched to metric/ISO about 1970 - proudly announcing so by attaching stickers or embossing markings - this certainly helped the worldwide acceptance.

Funny enough, non-metric measurement is *still* in broad use, plumbing is still based on inches ("Zoll" in Germany). And no one cares 8)


Funny enough, non-metric measurement is *still* in broad use

In fact, the other day I was wondering if there is a metric designation for the exact equivalent of what we call a ""1/4-20" bolt (i.e. 1/4" in diameter, with 20 threads per inch)??


1/4" = 6.35mm so I'm afraid there's no direct equivalent here AFAIK. And that's no surprise - there's no direct metric equivalent of medieval nuts and bolts either :-) (How should they have known until such a system of units was invented?)


Hi Dave,

No there are definitely no official metric equivalents to the US NC, NF and NPT thread sizes. Like was already pointed out. It would make for some weird standard sizes. The diameter would be weird enough but the pitch would be very weird. The metric equivalent of 1/4-20 would be 6.35-1.28mm .


Walter & Bruce,

Thanks for the info. Of course, I figured that there was likely to be no "exact" equivalent because the values would be weird. But I thought there might be some size that was close in name (a 6-13?) but well-understood to be essentially the equivalent. Sort of like a "2x4" in the US - the real size is nowhere close to 2 inches by 4 inches. (I think this was, or became, the subject of a thread quite some time ago.)


At the risk of being pedantic, 2 x 4 is the size of the unplanned board.


The matter is far easier with nuts and bolts here: you need one number only for the standard. What comes closest to what you mentioned is a thread called M6. If you need a nut with this thread, M6 is all you need to know. If you need a bolt of e.g. 50mm length, it's fully specified by M6 x 50 (plus an indication for the kind of head that bolt shall have). The outer diameter of said thread is almost 6mm, thus M6. See here for a little bit more:


Another bit of information I will store in my mind as "useful bits" :-). Thank for the clarification.


I think the abbreviated marking is more commonly found on Japanese components (also some wicked color markings on capacitors).


The .001 uF cap has split in half. In other words, on side of the conductor has de-laminated from the dielectric.

So even in this state, the calc should turn on if all other circuits are viable? It is in a very sorry state as it had a LOT of corrosion. It took quite a while in vinegar/water bath followed by alcohol to get it all off. Some of the long connecting pins have disintegrated. I am still not 100% confident the corrosion is cleared beneath some of the IC's

This may turn into one of my long cold Maine winter projects, much like my first 55.

Edited: 24 Sept 2012, 7:40 p.m.


On a second look, you might need the cap to get the motor to run correctly (both directions). That said, if the head is in the home position, the calculator portion would be enabled.

All I can say is wow, I've never seen that much corrosion damage. Even the connector pins have been dissolved. With that amount of metal gone, I'd be very concerned there is board conductor damage. If it were mine, I'd harvest the IC's from it and see any worked before getting too far down the drain in the time sink. Way, way more work there than what I would consider repairable. All the plastic will be very brittle as well so be careful bending anything.


This looks like a case for a complete rework. The corrosion might as well have affected solder points and PCB traces, completely dissolved traces are not uncommon. I'd suggest a really good cleaning, maybe even desoldering of the components and checking all parts and traces. Quite a bit of work, I agree, but the 19C is worth it:)


The plastic is actually in very good shape.

Does anyone have recommendations on how to test each IC?


I don't think it's necessary to test them (and there probably is no way to test them standalone). Be sure to replace the potentiometers though, might be a good idea to just replace all parts (besides the ICs) and cables.


Very nice pictures, interesting case!
If you allow me a simple advice, I suggest you, before any futher repair job or investigation, a vinegrade wash of the calculator, better to remove first any unwanted guest!

Edited: 25 Sept 2012, 5:36 p.m.


I have had some cheap non-HP calculators in a similar condition - had to repair tracks etc., but never so bad that the capacitors delaminated. Let's hope the IC's work, everything else is pretty much replaceable. Good luck.


Can not help but say: WOWWWW!!!

Agreed, save the IC's, take *everything* apart and rebuilt the whole circuit from scratch; probably the PCB too.

In my opinion this is more a reconstruction job than a repair.

Look forward to seeing the finished pictures.

Best of luck and a bit of *envy* (these are the sort of challenges I love the most... ;-)


Possibly Related Threads…
Thread Author Replies Views Last Post
  Hp-67 card reader: tantalum capacitor identification Ignacio S├ínchez 8 3,269 06-02-2013, 07:26 PM
Last Post: aj04062
  Looking for scans of a HP-19c Box ledudu 5 2,170 01-20-2013, 06:13 AM
Last Post: Lode
  Craigslist find and fix - HP 19C Matt Fegenbush 12 3,749 08-13-2012, 02:57 PM
Last Post: Matt Fegenbush
  Original DM-15CC batch capacitor patch Neil Hamilton (Ottawa) 2 1,379 07-19-2012, 08:26 PM
Last Post: Gerson W. Barbosa
  WP-34s: Crystal and capacitor instructions? W. Bruce Maguire II 2 1,320 05-03-2012, 02:04 PM
Last Post: W. Bruce Maguire II
  HP-19C with a strange S/N Ernst Fueloep 1 1,035 10-10-2011, 02:13 PM
Last Post: Juergen Keller
  HP-19c battery pack replacement (again...) Xavier A. (Brazil) 8 2,754 09-22-2011, 04:41 PM
Last Post: Xavier A. (Brazil)
  HP 19C continuing!!! Geoff Quickfall 3 1,486 07-28-2011, 03:08 PM
Last Post: Mark Henderson
  HP 19C horror show! Geoff Quickfall 17 4,126 06-14-2011, 03:59 PM
Last Post: Ernst Fueloep
  Help! HP-19C printer gone mad - fixed :-) Philippe Lasnier 1 1,038 04-25-2011, 10:08 PM
Last Post: Katie Wasserman

Forum Jump: